{
\def\sym#1{\ifmmode^{#1}\else\(^{#1}\)\fi}
\begin{tabular}{l*{4}{c}}
\hline\hline
            &\multicolumn{1}{c}{(1)}&\multicolumn{1}{c}{(2)}&\multicolumn{1}{c}{(3)}&\multicolumn{1}{c}{(4)}\\
            &\multicolumn{1}{c}{2003-2019}&\multicolumn{1}{c}{2003-2019}&\multicolumn{1}{c}{2004-2019}&\multicolumn{1}{c}{2004-2015}\\
\hline
2019        &      -15.92\sym{***}&      -23.03\sym{***}&      -17.20\sym{***}&      -23.86\sym{***}\\
            &     (0.347)         &     (0.591)         &     (0.420)         &     (0.606)         \\
[1em]
Treatment (absence of booth)&      -5.028\sym{***}&      -4.079\sym{***}&      -3.748\sym{***}&      -3.748\sym{***}\\
            &     (0.856)         &     (0.871)         &     (0.888)         &     (0.890)         \\
[1em]
2009-11     &                     &      -1.063\sym{*}  &                     &                     \\
            &                     &     (0.580)         &                     &                     \\
[1em]
2014-15     &                     &      -17.44\sym{***}&                     &                     \\
            &                     &     (0.586)         &                     &                     \\
[1em]
2009-11     &                     &                     &                     &      -1.809\sym{***}\\
            &                     &                     &                     &     (0.467)         \\
[1em]
2014-15     &                     &                     &                     &      -18.18\sym{***}\\
            &                     &                     &                     &     (0.501)         \\
\hline
\(N\)       &         420         &         420         &         448         &         448         \\
\hline\hline
\multicolumn{5}{l}{\footnotesize Standard errors in parentheses}\\
\multicolumn{5}{l}{\footnotesize All models include voting station * election type fixed effects}\\
\multicolumn{5}{l}{\footnotesize Standard errors are clustered by voting station * election type}\\
\multicolumn{5}{l}{\footnotesize \sym{*} \(p<0.10\), \sym{**} \(p<0.05\), \sym{***} \(p<0.01\)}\\
\end{tabular}
}
